• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 689 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1298R: ADS1298R Internal Test signals problem

Tsvetan Mudrov
Prodigy 220 points
Part Number: ADS1298R
Other Parts Discussed in Thread: ADS1298

Hello. We have a device with ADS1298R in the following configuration:

AVDD = 3.3V

AVSS = GND(0V)

DVDD = 3.3V

VREF = 2.4V (Internal)

Pace buffer = OFF

Vref buffer = ON

Oscillator - external, 2.048Mhz.

I have a trouble to use "internal test" signals. When I select input multiplexer to Test Signal path on a given channel - the result is not correct. It is a rectangle waveform but the amplitude is about 50uV (instead of 2mV). I have tried a lot of combinations with registers without success.

I have a trouble to read on-chip temperature also. In fact all positions of input multiplexer are not working correctly, except of "Normal Electrode" which is working just fine - I'm able to read correctly what I have on the analog input of the chip.

After a days without success, I bought a development kit. In this way a have a possibility to experiment with pin connections, which is impossible on our board (due to BGA package).

What I found is - when I connect TESTP_PACE_OUT1 and TESTP_PACE_OUT2 pins to AVDD - the "internal test" signal amplitude becomes wrong. The temperature and all other internal measurement channels are also wrong. On my board these pins are connected to AVDD, as recommended by data sheet (see below).


But:

1. On table "Pin Function: NFBGA Package" on page 7 in data sheet - unused pins must be connected to AVDD and it is explicative noted exactly for TESTP_PACE_OUTx pins. This is what we have done on our PCB.

2. Due to the "Figure 25. Input Multiplexer Block for One Channel" - When INT_TEST = 1, the TESTP_PACE_OUTx pins are disconnected from the internal multiplexer, so normally the voltage on the pin should not affect the internal measurements (test, temperature, vdd and etc.).

3. The PACE buffer is OFF, so it should not affect the internal measurements.

So what I'm doing wrong? Is it possible that connection of TESTP_PACE_OUTx pins to AVDD affects measurement of internal signals somehow?

Please, I need an urgent answer, since we have a serie of hundreds of boards which are mounted and waiting for firmware implementation of internal tests signals and on chip temperature measurement.

  • 0
    TI__Mastermind 38745 points
    Hello Tsvetan,

    Unfortunately this is an error in the datasheet we just discovered. You should float the TESTP_PACE_OUT1 and TESTN_PACE_OUT2 inputs. Is that possible in your design? Have you tied them high through a resistor or directly to AVDD?

    Regards,
    Brian Pisani
  • 0 in reply to Brian Pisani
    Prodigy 220 points

    Hello.

    I have strictly follow the data sheet. The both pins are connected to power plane under the BGA package. It's impossible to disconnect them.

    So us, I understand, there is no way to use internal channels of the input multiplexer in my configuration. Is this correct?

  • 0 in reply to Tsvetan Mudrov
    TI__Mastermind 38745 points

    Hello Tsvetan,

    Using the EVM, I was able to measure the test signal accurately when the TESTP/N_PACE_OUTx signals tied to AVDD when I enabled the pace amplifiers. The caveat is that the pace amplifiers will be shorted to AVDD so the ADS1298 will consume the extra short circuit current (specified in the datasheet as 490 uA per amplifier if AVDD - AVSS = 5V or 270 uA if AVDD - AVSS = 3V). Is this a feasible workaround?

    Brian

  • 0 in reply to Brian Pisani
    Prodigy 220 points

    I have tried on my EVM module - seems to works. When the power of Pace amplifier is turned ON, I'm able to read correctly test signals, even if both pins are connected to AVDD. I will try on real PCB later and will report you after that.

    It's a strange solution, can you explain a bit - how it works? Due to the schematics of the input multiplexer from the data sheet, such a solution looks confusing.

    Can you guarantee the safety of the chip and entire PCB? The additional consumption seems fine for our design.

  • 0 in reply to Tsvetan Mudrov
    TI__Mastermind 38745 points
    Hey Tsvetan,

    It's actually a combination of a design bug and a datasheet error. The mux diagrams in the datasheet are how it is supposed to work, but the device was implemented with a quirk that makes it so turning the pace amps on is the only thing that fully disconnects the test signal from the pins.

    Technically I cannot "guarantee" anything that is not in the datasheet, but I spoke with the silicon designer about this and he says it should not be a problem to short those pins. Hopefully that works for your team.

    Regards,
    Brian Pisani
  • 0 in reply to Brian Pisani
    Prodigy 220 points

    Yes I see.

    Ok, I will implement it in our firmware. The test procedure is relatively short, so I hope it will not be risky for the hardware. So according to me - we have a workaround.

    Thank you!

    P.S. Anyway - It will be great if you correct the datasheet in a reasonable time. It's so strange that I'm the first one which falls in that hole :)